All animals used for sample isolation were handled in accordance with the National Institute of Animal Sciences (NIAS) (Hanoi, Vietnam) care and management guidelines. Five milliliter (5 mL) of blood was collected with anticoagulant from jugular vein of the I pigs reared in the Thanh Hoa province (Vietnam) according to I pig breed conservation project of NIAS. Total DNA extraction was performed by standard phenol-chloroform method [10].

The full sequence of mtDNA was amplified by using 30 primer pairs into 30 overlap segments (Table 1). Each 25 μL polymerase chain reaction (PCR) reaction consist of 12.5 μL GoTaq Green Master Mix, 1.0 μL DNA template, 0.5 μL of each primer (10 ppm), and 10.5 μL sterile deionized water. The PCR thermocycler was an initial denaturation at 94°C for 5 min, followed by 25 cycles of 94°C for 30 s, 50°C to 55°C based on the primer sequences for 30 s, and 72°C for 50 min, with a final extension at 72°C for 10 min. The product sequence length was checked by 1% agarose gel electrophoresis. PCR products were purified by QIAquick PCR Purification Kit (Qiagen, Hilden, Germany).

The PCR products were sequenced using ABI3500 DNA sequencer. The DNA sequences were assembled using DNA Dragon v1.6.0 software and exported to FASTA file. The complete sequence was annotated using MITOS and DOGMA. The mitochondrial circle structure was built using GenomeVx and the tRNAs structure was downloaded from online report of MITOS analysis result. The C, G, A, T frequencies and A+T-content of the entire mitochondrial genome were calculated using DAMBE v6.3.17 [11]. After that, skewness for each protein coding gene was evaluated using the following formulas: GC skew = (G−C)/(G+C) and AT skew = (A−T)/(A+T).

The D loop and full sequence of I pig were aligned with other pig sequences using MUSCLE algorithm [12] of MEGA6 [13]. The best model for full mitochondrial genome group was HKY+G and TN93+I. We used Bayesian phylogenetic analysis in BEAST v1.8.3 [14] with Yule process and MCMC 10000000 to calculate the alignment data and Figure Tree v1.4.2 to build the phylogenetic trees.

The total length of I pig mitochondrial genome was 16,731 base pairs (bp) (Figure 1). Similar to other Sus scrofa mitogenomes [15,16], there were 2 ribosome RNA genes, 13 protein coding genes, 22 transfer RNA genes and a non-coding D loop region. The base composition was 34.66% for A, 26.24% for C, 13.35%, for G and 25.75% for T, the A+T content was 60.41%. The D loop region was 1,295 bp in length and located between tRNA-Phe and tRNA-Pro. In addition, there were 12 other non-coding spaces in the length of 1 to 32 bp and 8 overlaps in the length of 1 to 43 bp among all genes. The length of 12S and 16S rRNA was 960 and 1,570 bp, respectively. Twenty two transfer RNA genes were range from 59 bp to 75 bp in size. The total length of 13 protein coding genes was 11,413 bp ranging from 204 bp to 1,821 bp. In which, four genes (cytochrome c oxidase subunits [COX] 2, COX3, nicotinamide adenine dinucleotide dehydrogenase subunits [ND] 3, and ND4) had incomplete stop codon T--, that presumptively formed a complete stop codon by post-transcriptional polyadenylation. The start and stop of all coding genes were same as other pig breeds. The GC and AT skews showed the nucleotide composition bias in the mitogenome sequence. All protein and rRNA coding genes located in H strand exhibited negative GC skew and positive AT skew, this reflected the bias of Cytosine and Adenine. In the H strand, the most Cytosine bias was adenosine triphosphatase subunits (ATPase) 8 (−0.55), the less Cytosine bias was 16S rRNA (−0.11), whereas 12S rRNA gene had the most Adenine bias and COX1 gene had Cytosine content equal Guanine. The only one gene coded in L strand was ND6 had positive GC skew (0.56) and negative AT skew (−0.35) (Table 2). Because the tandem repeat motif “CGTGCGTACA” could cause heteroplasmy [17], we removed them out the mitogenome sequence before analysis. The component of I pig full sequence after remove motif was 34.83%, 26.19%, 13.15%, and 25.82% for A, C, G, and T, respectively. Compared the AT content, GC and AT skew between I pig sequence and other pig breeds, the range of AT content, GC skew and AT skew of published full sequence were (60.19; 60.82), (−0.34; −0.30) and (0.12; 0.15), respectively. The range of AT content, GC skew, and AT skew of published D loop sequence were (60.73; 61.80), (−0.38; −0.31); (0.15; 0.19), respectively. We found that the AT content of I pig D loop (60.09%) was smaller than other pigs, whereas its full sequence (60.65%) was larger than Bihu pig (60.19%). The D loop of I pig also had less Cytosine bias than most pig breeds, excepted WB Yunnan (−0.31). GC and AT skew of full genome and AT skew of D loop did not have too much different (Table 3).

The secondary structure of tRNAs was predicted by MITOS online software (Figure 2). Twenty one tRNAs excepted for tRNA Ser-1 had the typical cloverleaf structure. The tRNA Ser-1 dihydrouridine arm did not form a stable structure.

We built the phylogenetic trees of the I pig full sequence and D loop sequence with the published mitogenomes of pig breeds from three different geographical regions: European Countries, Mekong region and China. In both phylogenetic trees, most of the pigs divided into two clades: Asian clade and European clade, the genetic distances between them in full sequence and D loop sequence phylogenetic trees were 0.008 and 0.0137, respectively.

The twenty-seven public reference sequences were used to build the full sequence phylogenetic tree (Figure 3). There were two large groups in the Asian clade. Although the Berkshire and WB Korea were in Asian clade they felt outside of both large groups, the distance from WB Korea was 0.0045 and Berkshire was further (0.0053). Genetic distance between two groups was 0.0004. The I pig, Banna mini pig, and WB Vietnam and WB Hainan clustered a group. I pig was a sister taxon with the Banna mini (0.0002), close was the WB Vietnam (0.0016) while the WB Hainan had most distant relationship (0.0022).

There were 33 sequences in the D loop phylogenetic tree and this tree showed the different distribution (Figure 4). The WB Japan and WB Korea joined to the group of I pig. Therefore, this group could be separated into two subgroups: I pig and Banna mini continued to be cluster (0.0006) and WB Vietnam, WB Korea and WB Japan were another subgroup. The genetic distance between the two subgroups was 0.0065. WB Hainan felt out of this group, it and Taoyuan pig became sister taxons (0.0014). The genetic distance between the two big groups of the Asian clade was 0.0057.